How archery studies measure aiming

Three measures under one name

When a study says it measured aiming, it may have measured one of three things: where the bow points on the target, obtained with an optical system mounted on the riser; how much the bow accelerates and trembles, which accelerometers pick up; or the timing of the shot, meaning when the clicker falls, when the arrow leaves and how far the draw lengthens in the final seconds. Wearable sensors have a section of their own further down, and the findings on bow stability, in chronological order, are in the line on bow stability and aiming time; this page looks at the instruments and how far their figures can be combined.

The optical system: where the bow points

The Noptel ST-2000 was designed for rifle shooting: an infrared unit reads its own alignment from reflecting prisms fixed around the centre of the target. For the German juniors studied in 2006 it was adapted to fit the riser in place of the lower stabiliser, and it recorded the aim point 100 times a second. From the trace the authors derive the hold box, a rectangle built on the horizontal and vertical standard deviations and measured in square score points, which has the advantage of showing how far the aim moves in rings rather than in degrees.

Across seven archers and 66 arrows each, the hold box shrinks in every archer as the shot approaches, and in the final second it is smaller on arrows scoring ten than on those scoring eight or less: this is the basis of the entry on the aim settling in the final second. Between archers, however, the average size of the hold box does not show who shoots better: the highest mean, 9.31 points, goes with 0.315, while the smallest hold box, 0.295, goes with 9.18. The same instrument, which weighs about 160 grams, was later set against two video systems on the arrows of three archers from the German women's squad. Since it takes the place of the lower stabiliser, though, the bow being measured is not quite in its competition set-up.

The accelerometer: how much the bow moves

An accelerometer on the bow has no idea where the arrow is pointing: it measures how the bow's velocity changes, and indices of stability or tremor are derived from that. In exchange it is cheap and portable. In the eleven Japanese university archers, a 25-gram sensor sampling at 25 Hz takes as tremor the standard deviation of acceleration from 1.2 to 0.2 seconds before the release. The more the bow trembles, the lower the total score, with R = -0.79; on a single arrow, though, no significant correlation appears.

Looking inside the tremor means sampling faster and breaking the signal down into frequencies. With three accelerometers on the sight at 500 Hz, a Japanese engineering group found two peaks near 8 and 14 Hz, but measured on seated people who are not described as archers: this is where the entry on bow tremor at two frequencies starts. In sixteen young archers, with a sensor at 250 Hz, the spectrum is split into a slow band between 1 and 10 Hz and a fast one between 10 and 20, but the boundary is the authors' choice, and with a different split the same movement changes name, as the entry on slow sway and tremor shows. A 2020 study filters out everything above 5 Hz as vibration of the bow, which is the band the work on the sight calls tremor.

The task being recorded matters too. In the side-plank study the tremor comes from twenty university students with no archery experience, holding a twenty-pound bow at aim for fifteen seconds; in the ten Turkish traditional archers the accelerometers give jerk, which describes how smooth the movement is and not where the arrow goes.

Timing: clicker, release and sound

Some of these sensors mark moments in time. In 1987 a single-axis accelerometer on the riser, at 2,500 Hz, with the clicker and a short length of string used as switches, followed the bow's sideways displacement over the first thirty milliseconds, and it is the measurement behind the entry on stabilisers and bow movement. In 2006 a strain gauge glued to the clicker reconstructed draw length during the final pull: the archer with the highest mean drew the arrow back by nearly 5 mm on every shot, steadily and in much the same way each time.

The cheapest signal is sound. With a microphone half a metre behind seven archers from the German national teams, the click of the clicker, the thud of the limbs at the end of the power stroke and the impact on the target can be told apart. The time from clicker to shot ranged from 144.40 to 175.96 milliseconds depending on the archer, and its variability from one arrow to the next explained on its own 66% of the differences in mean score between archers: the entry on consistency of release timing rests on this, and on seven people.

Within one archer or between archers

The main trap is not technical. An index that separates better arrows from worse ones within the same archer may say nothing when different archers are compared: the 2006 hold box works in the first sense and not in the second, and the tremor measured by accelerometer in 2021 follows the score of the round and not that of the single arrow. In the four young Korean archers wearing nine inertial sensors, a model can be found for each one, but it differs for each and explains between 27.9 and 72.0% of the variability in score. The entry on movement parameters and score brings this together, and it points the same way as the entry on a single correct technique for the elite.

On top of this, every study measures a different quantity and calls it stability: the area covered by the aim, the amplitude of a sway, the standard deviation of an acceleration. That is why the entry on bow stability while aiming is contested, and whether the aim moves more horizontally than vertically depends partly on the instrument used to look at it.

What they can tell an archer

For an archer or a coach, these instruments are at their best when the comparison is with oneself: the same sensor, in the same place, with the same processing, over many arrows from the same archer. Under those conditions the settling of the aim in the final second, the regularity of the final pull and the consistency of the time after the clicker can be followed over time. Using another study's figure as a threshold is less secure, because the instrument, sampling rate, time window and unit all change. Likewise, the calculation by which a sight one metre from the eye, moving 1.74 millimetres, sends the arrow 122 millimetres further out at 70 metres is geometry, as the entry on millimetres at aim and centimetres on the target explains, not a movement recorded on an archer.

How design and sample are weighed is set out in How we read a study; for the archer, tremor is covered in the article on bow-arm tremor, and timing in those on aiming time and the final pull.

@misc{campagna2026_enresearchmethodsmeasuringaimsensors,
  author = {Campagna, Matteo},
  title = {How archery studies measure aiming},
  year = {2026},
  howpublished = {Research, matteocampagna.com},
  url = {https://matteocampagna.com/en/research/methods/measuring-aim-sensors},
  note = {accessed 2026-10-16}
}